search

cheng052 / BRNet

(CVPR 2021) Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds
Apache License 2.0
89 stars 16 forks source link

Collaboration with MMDetection3D #4

Open Tai-Wang opened 3 years ago

Tai-Wang commented 3 years ago

Hi, developers of BRNet,

Excited to find your project is reimplemented based on mmdet3d. It's good news that this framework benefits this community in terms of easier usage and development of research ideas.

I believe your work could bring great insights to this field, so I am wondering whether you have the willingness to contribute an implementation to MMDetection3D. I believe it could increase the impact of your work and also leave a mark on a fair and comprehensive benchmark.

We welcome any kind of contributions, so please feel free to ask us for help needed from our team. Thanks.

On behalf of the MMDet3D Development Team

Best, Tai

cheng052 commented 3 years ago

Hi, Tai,

I am willing to contribute an implementation to MMDetection3D, and could you please give some guidance on how to do this.

Best, Bowen

Tai-Wang commented 3 years ago

Great! After a brief review of the difference of this repo with mmdet3d, I think the basic structure is ok.

In my opinion, the main changes are focused on a class agnostic VoteHead, an RoI head for back-tracing and revisiting representative points and further detection as well as specific configs for your model, right? Then we can just split these important modules into different small Pull Requests (maybe about 3 PRs for above components and 1 for benchmark (readme page)) and clean them one by one.

So next, you just need to migrate the related code to mmdet3d and refine some docstrings/code styles. We can help you review your codes. Here is a basic guidance to leverage the pre-commit tool to refine your codes and contribute to open-mmlab codebases. You can also refer to our codes for the basic style and code standards. Let's get together to make the codes clearer and easier to use~

cheng052 commented 3 years ago

Thanks for your helpful guidance!

I will begin this work once I have time. I have a lot of work recently, sorry for the late contribution.

Keep in touch.